platform/kernel/linux-rpi.git
3 years agoselftests: net: devlink_port_split.py: skip the test if no devlink device
Po-Hsu Lin [Thu, 20 May 2021 10:49:54 +0000 (18:49 +0800)]
selftests: net: devlink_port_split.py: skip the test if no devlink device

When there is no devlink device, the following command will return:
  $ devlink -j dev show
  {dev:{}}

This will cause IndexError when trying to access the first element
in dev of this json dataset. Use the kselftest framework skip code
to skip this test in this case.

Example output with this change:
  # selftests: net: devlink_port_split.py
  # no devlink device was found, test skipped
  ok 7 selftests: net: devlink_port_split.py # SKIP

Link: https://bugs.launchpad.net/bugs/1928889
Signed-off-by: Po-Hsu Lin <po-hsu.lin@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoibmvnic: remove default label from to_string switch
Michal Suchanek [Thu, 20 May 2021 06:50:34 +0000 (08:50 +0200)]
ibmvnic: remove default label from to_string switch

This way the compiler warns when a new value is added to the enum but
not to the string translation like:

drivers/net/ethernet/ibm/ibmvnic.c: In function 'adapter_state_to_string':
drivers/net/ethernet/ibm/ibmvnic.c:832:2: warning: enumeration value 'VNIC_FOOBAR' not handled in switch [-Wswitch]
  switch (state) {
  ^~~~~~
drivers/net/ethernet/ibm/ibmvnic.c: In function 'reset_reason_to_string':
drivers/net/ethernet/ibm/ibmvnic.c:1935:2: warning: enumeration value 'VNIC_RESET_FOOBAR' not handled in switch [-Wswitch]
  switch (reason) {
  ^~~~~~

Signed-off-by: Michal Suchanek <msuchanek@suse.de>
Acked-by: Lijun Pan <lijunp213@gmail.com>
Link: https://lore.kernel.org/netdev/CAOhMmr701LecfuNM+EozqbiTxFvDiXjFdY2aYeKJYaXq9kqVDg@mail.gmail.com/
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoNFC: st21nfca: remove unnecessary variable and labels
wengjianfeng [Thu, 20 May 2021 01:05:50 +0000 (09:05 +0800)]
NFC: st21nfca: remove unnecessary variable and labels

assign vlue (EIO/EPROTO) to variable r, and goto exit label,
but just return r follow exit label, so we delete exit label,
and just replace with return sentence.

Signed-off-by: wengjianfeng <wengjianfeng@yulong.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'bond-cleanups'
David S. Miller [Thu, 20 May 2021 22:43:25 +0000 (15:43 -0700)]
Merge branch 'bond-cleanups'

Guangbin Huang says:

====================
net: bonding: clean up some code style issues

This patchset cleans up some code style issues.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: bonding: use tabs instead of space for code indent
Yufeng Mo [Thu, 20 May 2021 06:18:35 +0000 (14:18 +0800)]
net: bonding: use tabs instead of space for code indent

Code indent should use tabs where possible, so
use tabs instead of space for code indent.

Signed-off-by: Yufeng Mo <moyufeng@huawei.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: bonding: remove unnecessary braces
Yufeng Mo [Thu, 20 May 2021 06:18:34 +0000 (14:18 +0800)]
net: bonding: remove unnecessary braces

Braces {} are not necessary for single statement blocks,
so remove these braces {}.

Signed-off-by: Yufeng Mo <moyufeng@huawei.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: bonding: fix code indent for conditional statements
Yufeng Mo [Thu, 20 May 2021 06:18:33 +0000 (14:18 +0800)]
net: bonding: fix code indent for conditional statements

Fix incorrect code indent for conditional statements.

Signed-off-by: Yufeng Mo <moyufeng@huawei.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: bonding: add some required blank lines
Yufeng Mo [Thu, 20 May 2021 06:18:32 +0000 (14:18 +0800)]
net: bonding: add some required blank lines

Add some blank lines after declarations as required.

Signed-off-by: Yufeng Mo <moyufeng@huawei.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue
David S. Miller [Thu, 20 May 2021 22:13:28 +0000 (15:13 -0700)]
Merge branch '1GbE' of git://git./linux/kernel/git/tnguy/next-queue

Tony Nguyen says:

====================
1GbE Intel Wired LAN Driver Updates 2021-05-20

This series contains updates to igc driver only.

Andre Guedes says:

This series adds AF_XDP zero-copy feature to igc driver.

The initial patches do some code refactoring, preparing the code base to
land the AF_XDP zero-copy feature, avoiding code duplications. The last
patches of the series are the ones implementing the feature.

The last patch which indeed implements AF_XDP zero-copy support was
originally way too lengthy so, for the sake of code review, I broke it
up into two patches: one adding support for the RX functionality and the
other one adding TX support.
---
v2:
Patch 8/9 - "igc: Enable RX via AF_XDP zero-copy"
 * In XDP_PASS flow, copy metadata too into the skb.
 * When HW timestamp is added by the NIC, after copying it into
   a local variable, update xdp_buff->data_meta so that
   metadata length when XDP program is called 0.
 * In igc_xdp_enable_pool(), call xsk_pool_dma_unmap() on
   failure.

Known issues:
 When an XDP application is running in Tx-Only mode with Zero-Copy
 enabled, it is not expected to add the frames to the fill-queue. I have
 noticed the following two issues in this scenario:
 - If XDP_USE_NEED_WAKEUP flag is not set by application, igc_poll()
   will go into infinite loop because the buffer allocation resulting
   in igc_clean_rx_irq_zc() indicating that all work is not done and NAPI
   should keep polling. This does not occur if XDP_USE_NEED_WAKEUP
   flag is set.
 - Since there are no buffers allocated by userspace for the fill
   queue, there is no memory allocated for the NIC to copy the data
   to. If the packet received is destined to the hardware queue where
   XDP application is running, no packets are received even on other
   queues.
 Both these issues can be mitigated by adding a few frames to the
 fill queue. The second issue can also be mitigated by making sure no
 packets are being received on the hardware queue where Rx is running.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'net-leading-spaces'
David S. Miller [Thu, 20 May 2021 22:10:57 +0000 (15:10 -0700)]
Merge branch 'net-leading-spaces'

Hui Tang says:

====================
net: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

        $ find . -name '*.[ch]' | xargs sed -r -i 's/^[ ]+\t/\t/'
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomii: remove leading spaces before tabs
Hui Tang [Thu, 20 May 2021 03:47:54 +0000 (11:47 +0800)]
mii: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running
the following commard:

    $ find . -name '*.[ch]' | xargs sed -r -i 's/^[ ]+\t/\t/'

Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoifb: remove leading spaces before tabs
Hui Tang [Thu, 20 May 2021 03:47:53 +0000 (11:47 +0800)]
ifb: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running
the following commard:

    $ find . -name '*.[ch]' | xargs sed -r -i 's/^[ ]+\t/\t/'

Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: appletalk: remove leading spaces before tabs
Hui Tang [Thu, 20 May 2021 03:47:52 +0000 (11:47 +0800)]
net: appletalk: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running
the following commard:

    $ find . -name '*.[ch]' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: fddi: skfp: remove leading spaces before tabs
Hui Tang [Thu, 20 May 2021 03:47:51 +0000 (11:47 +0800)]
net: fddi: skfp: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running
the following commard:

    $ find . -name '*.[ch]' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hamradio: remove leading spaces before tabs
Hui Tang [Thu, 20 May 2021 03:47:50 +0000 (11:47 +0800)]
net: hamradio: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running
the following commard:

    $ find . -name '*.[ch]' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Jiri Slaby <jirislaby@kernel.org>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: ppp: remove leading spaces before tabs
Hui Tang [Thu, 20 May 2021 03:47:49 +0000 (11:47 +0800)]
net: ppp: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running
the following commard:

    $ find . -name '*.[ch]' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Tom Parkin <tparkin@katalix.com>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: slip: remove leading spaces before tabs
Hui Tang [Thu, 20 May 2021 03:47:48 +0000 (11:47 +0800)]
net: slip: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running
the following commard:

    $ find . -name '*.[ch]' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Masahiro Yamada <masahiroy@kernel.org>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: usb: remove leading spaces before tabs
Hui Tang [Thu, 20 May 2021 03:47:47 +0000 (11:47 +0800)]
net: usb: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running
the following commard:

    $ find . -name '*.[ch]' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Hayes Wang <hayeswang@realtek.com>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: wan: remove leading spaces before tabs
Hui Tang [Thu, 20 May 2021 03:47:46 +0000 (11:47 +0800)]
net: wan: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running
the following commard:

$ find . -name '*.[ch]' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Xie He <xie.he.0141@gmail.com>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'hns3-debugfs'
David S. Miller [Thu, 20 May 2021 22:01:04 +0000 (15:01 -0700)]
Merge branch 'hns3-debugfs'

Huazhong Tan says:

====================
net: hns3: refactor some debugfs commands

This series refactors the debugfs command to the new
process and removes the useless debugfs file node cmd
for the HNS3 ethernet driver.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: remove the useless debugfs file node cmd
Yufeng Mo [Thu, 20 May 2021 02:21:44 +0000 (10:21 +0800)]
net: hns3: remove the useless debugfs file node cmd

Currently, all debugfs commands have been reconstructed, and the
debugfs file node cmd is useless. So remove this debugfs file node.

Signed-off-by: Yufeng Mo <moyufeng@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor dump serv info of debugfs
Yufeng Mo [Thu, 20 May 2021 02:21:43 +0000 (10:21 +0800)]
net: hns3: refactor dump serv info of debugfs

Currently, the debugfs command for serv info is implemented by
"echo xxxx > cmd", and record the inforamtion in dmesg. It's
unnecessary and heavy. To improve it, create a single file
"serv_info" for it, and query it by command "cat serv_info",
return the result to userspace, rather than record in dmesg.

The display style is below:
$ cat service_task_info
local_clock: [  114.203321]
delta: 784(ms)
last_service_task_processed: 4294918512(jiffies)
last_service_task_cnt: 4

Signed-off-by: Yufeng Mo <moyufeng@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor dump mac tnl status of debugfs
Jiaran Zhang [Thu, 20 May 2021 02:21:42 +0000 (10:21 +0800)]
net: hns3: refactor dump mac tnl status of debugfs

Currently, the debugfs command for dump mac tnl status is
implemented by "echo xxxx > cmd", and record the information
in dmesg. It's unnecessary and heavy. To improve it, create
a single file "mac_tnl_status" for it, and query it by command
"cat mac_tnl_status", return the result to userspace, rather
than record in dmesg.

The display style is below:
$ cat mac_tnl_status
Recently generated mac tnl interruption:
[0111204.175437] status = 0x30
[0154120.329912] status = 0x30

Signed-off-by: Jiaran Zhang <zhangjiaran@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor dump qs shaper of debugfs
Guangbin Huang [Thu, 20 May 2021 02:21:41 +0000 (10:21 +0800)]
net: hns3: refactor dump qs shaper of debugfs

Currently, user gets qset shaper parameters by implementing debugfs
command "echo dump qs shaper > cmd", this command will dump info in
dmesg. It's unnecessary and heavy.

As there is "tm_qset" file in tm directory for dump qset info, to
optimize these command, merge qset shaper parameters to tm_qset
file and use cat command to get them.

The display style is below:
$ cat tm_qset
ID    MAP_PRI  LINK_VLD  MODE  DWRR  IR_B  IR_U  IR_S  BS_B  BS_S  FLAG
0000     0        1      dwrr  100   150     7     0     5    20     0
0001     0        0        sp    0   150     7     0     5    20     0

Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor dump qos buf cfg of debugfs
Guangbin Huang [Thu, 20 May 2021 02:21:40 +0000 (10:21 +0800)]
net: hns3: refactor dump qos buf cfg of debugfs

Currently, user gets qos buffer config by implementing debugfs command
"echo dump qos buf cfg > cmd", this command will dump info in dmesg.
It's unnecessary and heavy.

To optimize it, create a single file "qos_buf_cfg" in tm directory
and use cat command to get info. It will return info to userspace,
rather than record in dmesg.

The display style is below:
$ cat qos_buf_cfg
tx_packet_buf_tc_0: 0x120
tx_packet_buf_tc_1: 0x120
tx_packet_buf_tc_2: 0x120
tx_packet_buf_tc_3: 0x120
tx_packet_buf_tc_4: 0x0
tx_packet_buf_tc_5: 0x0
tx_packet_buf_tc_6: 0x0
tx_packet_buf_tc_7: 0x0
......

Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor dump qos pri map of debugfs
Guangbin Huang [Thu, 20 May 2021 02:21:39 +0000 (10:21 +0800)]
net: hns3: refactor dump qos pri map of debugfs

Currently, user gets priority map by implementing debugfs command
"echo dump qos pri map > cmd", this command will dump info in dmesg.
It's unnecessary and heavy.

To optimize it, create a single file "qos_pri_map" in tm directory
and use cat command to get info. It will return info to userspace,
rather than record in dmesg.

The display style is below:
$ cat qos_pri_map
vlan_to_pri: 0
PRI    TC
0       0
1       1
2       2
3       3
4       0
5       1
6       2

Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor dump qos pause cfg of debugfs
Guangbin Huang [Thu, 20 May 2021 02:21:38 +0000 (10:21 +0800)]
net: hns3: refactor dump qos pause cfg of debugfs

Currently, user gets pause config by implementing debugfs command
"echo dump qos pause cfg > cmd", this command will dump info in dmesg.
It's unnecessary and heavy.

To optimize it, create a single file "qos_pause_cfg" in tm directory
and use cat command to get info. It will return info to userspace,
rather than record in dmesg.

The display style is below:
$ cat qos_pause_cfg
pause_trans_gap: 0x7f
pause_trans_time: 0xffff

Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor dump tc of debugfs
Guangbin Huang [Thu, 20 May 2021 02:21:37 +0000 (10:21 +0800)]
net: hns3: refactor dump tc of debugfs

Currently, user gets tc schedule info by implementing debugfs command
"echo dump tc > cmd", this command will dump info in dmesg. It's
unnecessary and heavy.

To optimize it, create a single file "tc_sch_info" and use cat command
to get info. It will return info to userspace, rather than record in
dmesg.

The display style is below:
$ cat tc_sch_info
enabled tc number: 4
weight_offset: 14
TC    MODE  WEIGHT
0     dwrr     25
1     dwrr     25
2     dwrr     25
3     dwrr     25
4     dwrr      0
5     dwrr      0
6     dwrr      0
7     dwrr      0

Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor dump tm of debugfs
Guangbin Huang [Thu, 20 May 2021 02:21:36 +0000 (10:21 +0800)]
net: hns3: refactor dump tm of debugfs

Currently, user gets some tm info by implementing debugfs command
"echo dump tm > cmd", this command will dump info in dmesg. It's
unnecessary and heavy.

In addition, the info of this command mixes info of qset, priority,
pg and port. Qset and priority have their own command to get info of
themself, so can remove info of qset and priority from this command.

To optimize it, create two new files "tm_pg", "tm_port" in tm directory
and use cat command to separately get info of pg and port.

The display style is below:
$ cat tm_pg
ID  PRI_MAP  MODE DWRR  C_IR_B  C_IR_U  C_IR_S  C_BS_B  C_BS_S ...
00   0x1f    dwrr  1       75       9       0      31      20  ...

$ cat tm_port
IR_B  IR_U  IR_S  BS_B  BS_S  FLAG  RATE(Mbps)
75     9     0    31    20    1     200000

Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor dump tm map of debugfs
Guangbin Huang [Thu, 20 May 2021 02:21:35 +0000 (10:21 +0800)]
net: hns3: refactor dump tm map of debugfs

Currently, the debugfs command for tm map is implemented by
"echo xxxx > cmd", and record the information in dmesg. It's
unnecessary and heavy. To improve it, create a single file
"tm_map" for it, and query it by command "cat tm_map",
return the result to userspace, rather than record in dmesg.

As user can't specify queue id in cat command, driver will return info
of all queue id.

The display style is below:
$ cat tm_map
queue_id   qset_id   pri_id   tc_id
0000         0000      00       00
INDEX | TM BP QSET MAPPING:
0000  | 00000000:00000000:00000000:00000000:00000000:00000000:00000000
0256  | 00000000:00000000:00000000:00000000:00000000:00000002:00000000
0512  | 00000000:00000000:00000000:00000004:00000000:00000000:00000000
0768  | 00000000:00000008:00000000:00000000:00000000:00000000:00000000

Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor dump fd tcam of debugfs
Hao Chen [Thu, 20 May 2021 02:21:34 +0000 (10:21 +0800)]
net: hns3: refactor dump fd tcam of debugfs

Currently, the debugfs command for fd tcam is implemented by
"echo xxxx > cmd", and record the information in dmesg. It's
unnecessary and heavy. To improve it, create a single file
"fd_tcam" for it, and query it by command "cat fd_tcam",
return the result to userspace, rather than record in dmesg.

The display style is below:
$ cat fd_tcam
read result tcam key x(31):
00000000
00000000
00000000
08000000
00000600
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
read result tcam key y(31):
00000000
00000000
00000000
f7ff0000
0000f900
00000000
00000000
00000000
00000000
00000000
00000000
00000000
0000fff8

Signed-off-by: Hao Chen <chenhao288@hisilicon.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor queue info of debugfs
Hao Chen [Thu, 20 May 2021 02:21:33 +0000 (10:21 +0800)]
net: hns3: refactor queue info of debugfs

Currently, the debugfs command for queue info is implemented by
"echo xxxx > cmd", and record the information in dmesg. It's
unnecessary and heavy. To improve it, create two files
"rx_queue_info" and "tx_queue_info" for it, and query it
by command "cat rx_queue_info" and "cat tx_queue_info",
return the result to userspace, rather than record in dmesg.

The display style is below:
$ cat rx_queue_info
QUEUE_ID  BD_NUM  BD_LEN  TAIL  HEAD  FBDNUM  PKTNUM   ...
0           0       0     0     0       0       0      ...
1           0       0     0     0       0       0      ...
2           0       0     0     0       0       0      ...

$ cat tx_queue_info
QUEUE_ID  BD_NUM  TC  TAIL  HEAD  FBDNUM  OFFSET  PKTNUM  ...
0           0     0     0     0       0       0        0  ...
1           0     0     0     0       0       0        0  ...
2           0     0     0     0       0       0        0  ...

Signed-off-by: Hao Chen <chenhao288@hisilicon.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor queue map of debugfs
Hao Chen [Thu, 20 May 2021 02:21:32 +0000 (10:21 +0800)]
net: hns3: refactor queue map of debugfs

Currently, the debugfs command for queue map is implemented by
"echo xxxx > cmd", and record the information in dmesg. It's
unnecessary and heavy. To improve it, create a single file
"queue_map" for it, and query it by command "cat queue_map",
return the result to userspace, rather than record in dmesg.

The display style is below:
$ cat queue_map
local_queue_id   global_queue_id   vector_id
0                0                 341

Signed-off-by: Hao Chen <chenhao288@hisilicon.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor dump reg dcb info of debugfs
Yufeng Mo [Thu, 20 May 2021 02:21:31 +0000 (10:21 +0800)]
net: hns3: refactor dump reg dcb info of debugfs

Currently, the debugfs command for reg dcb info is implemented by
"echo xxxx > cmd", and record the information in dmesg. It's
unnecessary and heavy. To improve it, create a single file
"dcb" for it, and query it by command "cat dcb",
return the result to userspace, rather than record in dmesg.

The display style is below:
$ cat dcb
qset_id  roce_qset_mask  nic_qset_mask  qset_shaping_pass  qset_bp_status
0000           0x1            0x1             0x1               0x0
0001           0x1            0x1             0x1               0x0
0002           0x1            0x1             0x1               0x0
0003           0x1            0x1             0x1               0x0
0004           0x1            0x1             0x1               0x0
0005           0x1            0x1             0x1               0x0
0006           0x1            0x1             0x1               0x0
0007           0x1            0x1             0x1               0x0
pri_id  pri_mask  pri_cshaping_pass  pri_pshaping_pass
000       0x1           0x0                0x1
001       0x1           0x0                0x0
002       0x1           0x0                0x0
003       0x1           0x0                0x0
004       0x1           0x0                0x0
005       0x1           0x0                0x0
006       0x1           0x0                0x0
007       0x1           0x0                0x0
pg_id  pg_mask  pg_cshaping_pass  pg_pshaping_pass
000      0x1           0x0               0x1

Signed-off-by: Yufeng Mo <moyufeng@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: hns3: refactor dump reg of debugfs
Yufeng Mo [Thu, 20 May 2021 02:21:30 +0000 (10:21 +0800)]
net: hns3: refactor dump reg of debugfs

Currently, the debugfs command for reg is implemented by
"echo xxxx > cmd", and record the information in dmesg. It's
unnecessary and heavy. To improve it, create some files
"bios_common/ssu/igu_egu/rpu/ncsi/rtc/ppp/rcb/tqp/mac" for it,
and query it by command "cat xxx", return the result to
userspace, rather than record in dmesg.

The display style is below:
$ cat bios_common
BP_CPU_STATE: 0x0
DFX_MSIX_INFO_NIC_0: 0xc000
DFX_MSIX_INFO_NIC_1: 0x0
DFX_MSIX_INFO_NIC_2: 0x0
DFX_MSIX_INFO_NIC_3: 0x0
DFX_MSIX_INFO_ROC_0: 0xc000
DFX_MSIX_INFO_ROC_1: 0x0
DFX_MSIX_INFO_ROC_2: 0x0
DFX_MSIX_INFO_ROC_3: 0x0

Signed-off-by: Yufeng Mo <moyufeng@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomISDN: Remove obsolete PIPELINE_DEBUG debugging information
Zhen Lei [Thu, 20 May 2021 02:14:11 +0000 (10:14 +0800)]
mISDN: Remove obsolete PIPELINE_DEBUG debugging information

As Leon Romanovsky's tips:
The definition of macro PIPELINE_DEBUG is commented more than 10 years ago
and can be seen as a dead code that should be removed.

Suggested-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoigc: Enable TX via AF_XDP zero-copy
Andre Guedes [Fri, 23 Apr 2021 06:25:55 +0000 (23:25 -0700)]
igc: Enable TX via AF_XDP zero-copy

Add support for transmitting packets via AF_XDP zero-copy mechanism.

The packet transmission itself is implemented by igc_xdp_xmit_zc() which
is called from igc_clean_tx_irq() when the ring has AF_XDP zero-copy
enabled. Likewise i40e and ice drivers, the transmission budget used is
the number of descriptors available on the ring.

A new tx buffer type is introduced to 'enum igc_tx_buffer_type' to
indicate the tx buffer uses memory from xsk pool so it can be properly
cleaned after transmission or when the ring is cleaned.

The I225 controller has only 4 Tx hardware queues so the main difference
between igc and other Intel drivers that support AF_XDP zero-copy is
that there is no tx ring dedicated exclusively to XDP. Instead, tx
rings are shared between the network stack and XDP, and netdev queue
lock is used to ensure mutual exclusion. This is the same approach
implemented to support XDP_TX and XDP_REDIRECT actions.

Signed-off-by: Andre Guedes <andre.guedes@intel.com>
Signed-off-by: Vedang Patel <vedang.patel@intel.com>
Signed-off-by: Jithu Joseph <jithu.joseph@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
3 years agoigc: Enable RX via AF_XDP zero-copy
Andre Guedes [Fri, 23 Apr 2021 06:25:54 +0000 (23:25 -0700)]
igc: Enable RX via AF_XDP zero-copy

Add support for receiving packets via AF_XDP zero-copy mechanism.

Add a new flag to 'enum igc_ring_flags_t' to indicate the ring has
AF_XDP zero-copy enabled so proper ring setup is carried out during ring
configuration in igc_configure_rx_ring().

RX buffers can now be allocated via the shared pages mechanism (default
behavior of the driver) or via xsk pool (when AF_XDP zero-copy is
enabled) so a union is added to the 'struct igc_rx_buffer' to cover both
cases.

When AF_XDP zero-copy is enabled, rx buffers are allocated from the xsk
pool using the new helper igc_alloc_rx_buffers_zc() which is the
counterpart of igc_alloc_rx_buffers().

Likewise other Intel drivers that support AF_XDP zero-copy, in igc we
have a dedicated path for cleaning up rx irqs when zero-copy is enabled.
This avoids adding too many checks within igc_clean_rx_irq(), resulting
in a more readable and efficient code since this function is called from
the hot-path of the driver.

Signed-off-by: Andre Guedes <andre.guedes@intel.com>
Signed-off-by: Vedang Patel <vedang.patel@intel.com>
Signed-off-by: Jithu Joseph <jithu.joseph@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
3 years agoigc: Replace IGC_TX_FLAGS_XDP flag by an enum
Andre Guedes [Fri, 23 Apr 2021 06:25:53 +0000 (23:25 -0700)]
igc: Replace IGC_TX_FLAGS_XDP flag by an enum

Up to this point, Tx buffers are associated with either a skb or a xdpf,
and the IGC_TX_FLAGS_XDP flag was enough to distinguish between these
two case. However, with upcoming patches that will add AF_XDP zero-copy
support, a third case will be introduced so this flag-based approach
won't fit well.

In preparation to land AF_XDP zero-copy support, replace the
IGC_TX_FLAGS_XDP flag by an enum which will be extended once zero-copy
support is introduced to the driver.

Signed-off-by: Andre Guedes <andre.guedes@intel.com>
Signed-off-by: Vedang Patel <vedang.patel@intel.com>
Signed-off-by: Jithu Joseph <jithu.joseph@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
3 years agoigc: Introduce igc_unmap_tx_buffer() helper
Andre Guedes [Fri, 23 Apr 2021 06:25:52 +0000 (23:25 -0700)]
igc: Introduce igc_unmap_tx_buffer() helper

In preparation for AF_XDP zero-copy support, encapsulate the code that
unmaps Tx buffers into its own local helper so we can reuse it, avoiding
code duplication.

Signed-off-by: Andre Guedes <andre.guedes@intel.com>
Signed-off-by: Vedang Patel <vedang.patel@intel.com>
Signed-off-by: Jithu Joseph <jithu.joseph@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
3 years agoigc: Introduce TX/RX stats helpers
Andre Guedes [Fri, 23 Apr 2021 06:25:51 +0000 (23:25 -0700)]
igc: Introduce TX/RX stats helpers

In preparation for AF_XDP zero-copy support, encapsulate the code that
updates the driver RX stats in its own local helper so it can be reused
in the zero-copy path. Likewise, encapsulate TX stats code as well.

Signed-off-by: Andre Guedes <andre.guedes@intel.com>
Signed-off-by: Vedang Patel <vedang.patel@intel.com>
Signed-off-by: Jithu Joseph <jithu.joseph@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
3 years agoigc: Refactor XDP rxq info registration
Andre Guedes [Fri, 23 Apr 2021 06:25:50 +0000 (23:25 -0700)]
igc: Refactor XDP rxq info registration

Refactor XDP rxq info registration code, preparing the driver for AF_XDP
zero-copy support which is added by upcoming patches.

Currently, xdp_rxq and memory model are both registered during RX
resource setup time by igc_xdp_register_rxq_info() helper. With AF_XDP,
we want to register the memory model later on while configuring the ring
because we will know which memory model type to register
(MEM_TYPE_PAGE_SHARED or MEM_TYPE_XSK_BUFF_POOL).

The helpers igc_xdp_register_rxq_info() and igc_xdp_unregister_rxq_
info() are not useful anymore so they are removed.

Signed-off-by: Andre Guedes <andre.guedes@intel.com>
Signed-off-by: Vedang Patel <vedang.patel@intel.com>
Signed-off-by: Jithu Joseph <jithu.joseph@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
3 years agoigc: Refactor igc_clean_rx_ring()
Andre Guedes [Fri, 23 Apr 2021 06:25:49 +0000 (23:25 -0700)]
igc: Refactor igc_clean_rx_ring()

Refactor igc_clean_rx_ring() helper, preparing the code for AF_XDP
zero-copy support which is added by upcoming patches.

The refactor consists of encapsulating page-shared specific code into
its own helper, leaving common code that will be shared by both
page-shared and xsk pool in igc_clean_rx_ring().

Signed-off-by: Andre Guedes <andre.guedes@intel.com>
Signed-off-by: Vedang Patel <vedang.patel@intel.com>
Signed-off-by: Jithu Joseph <jithu.joseph@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
3 years agoigc: Refactor __igc_xdp_run_prog()
Andre Guedes [Fri, 23 Apr 2021 06:25:48 +0000 (23:25 -0700)]
igc: Refactor __igc_xdp_run_prog()

Refactor __igc_xdp_run_prog() helper from igc_xdp_run_prog(),
preparing the code for AF_XDP zero-copy support which is added
by upcoming patches.

The existing igc_xdp_run_prog() caters to regular XDP rx path
which has to verify if bpf_prog is not NULL. Zero-copy
path assumes that bpf_prog is not NULL and hence this check is
not required. Therefore it makes sense to refactor the common
code into a helper function, to avoid code duplication.

Signed-off-by: Andre Guedes <andre.guedes@intel.com>
Signed-off-by: Vedang Patel <vedang.patel@intel.com>
Signed-off-by: Jithu Joseph <jithu.joseph@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
3 years agoigc: Move igc_xdp_is_enabled()
Andre Guedes [Fri, 23 Apr 2021 06:25:47 +0000 (23:25 -0700)]
igc: Move igc_xdp_is_enabled()

Move the helper igc_xdp_is_enabled() to igc_xdp.h so it can be reused in
igc_xdp.c by upcoming patches that will introduce AF_XDP zero-copy
support to the driver.

Signed-off-by: Andre Guedes <andre.guedes@intel.com>
Signed-off-by: Vedang Patel <vedang.patel@intel.com>
Signed-off-by: Jithu Joseph <jithu.joseph@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
3 years agoMerge branch 'mt7530-interrupt-support'
David S. Miller [Wed, 19 May 2021 20:27:42 +0000 (13:27 -0700)]
Merge branch 'mt7530-interrupt-support'

DENG Qingfang says:

====================
MT7530 interrupt support

Add support for MT7530 interrupt controller.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agostaging: mt7621-dts: enable MT7530 interrupt controller
DENG Qingfang [Wed, 19 May 2021 03:32:02 +0000 (11:32 +0800)]
staging: mt7621-dts: enable MT7530 interrupt controller

Enable MT7530 interrupt controller in the MT7621 SoC.

Signed-off-by: DENG Qingfang <dqfext@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agodt-bindings: net: dsa: add MT7530 interrupt controller binding
DENG Qingfang [Wed, 19 May 2021 03:32:01 +0000 (11:32 +0800)]
dt-bindings: net: dsa: add MT7530 interrupt controller binding

Add device tree binding to support MT7530 interrupt controller.

Signed-off-by: DENG Qingfang <dqfext@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: dsa: mt7530: add interrupt support
DENG Qingfang [Wed, 19 May 2021 03:32:00 +0000 (11:32 +0800)]
net: dsa: mt7530: add interrupt support

Add support for MT7530 interrupt controller to handle internal PHYs.
In order to assign an IRQ number to each PHY, the registration of MDIO bus
is also done in this driver.

Signed-off-by: DENG Qingfang <dqfext@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: phy: add MediaTek Gigabit Ethernet PHY driver
DENG Qingfang [Wed, 19 May 2021 03:31:59 +0000 (11:31 +0800)]
net: phy: add MediaTek Gigabit Ethernet PHY driver

Add support for MediaTek Gigabit Ethernet PHYs found in MT7530 and
MT7531 switches.
The initialization procedure is from the vendor driver, but due to lack
of documentation, the function of some register values remains unknown.

Signed-off-by: DENG Qingfang <dqfext@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: qrtr: ns: Fix error return code in qrtr_ns_init()
Wei Yongjun [Wed, 19 May 2021 15:58:52 +0000 (15:58 +0000)]
net: qrtr: ns: Fix error return code in qrtr_ns_init()

Fix to return a negative error code -ENOMEM from the error handling
case instead of 0, as done elsewhere in this function.

Fixes: c6e08d6251f3 ("net: qrtr: Allocate workqueue before kernel_bind")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: ethernet: ixp4xx: Fix return value check in ixp4xx_eth_probe()
Wei Yongjun [Wed, 19 May 2021 14:16:27 +0000 (14:16 +0000)]
net: ethernet: ixp4xx: Fix return value check in ixp4xx_eth_probe()

In case of error, the function mdiobus_get_phy() returns NULL
pointer not ERR_PTR(). The IS_ERR() test in the return value
check should be replaced with NULL test.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet/sched: cls_api: increase max_reclassify_loop
Davide Caratti [Wed, 19 May 2021 13:17:21 +0000 (15:17 +0200)]
net/sched: cls_api: increase max_reclassify_loop

modern userspace applications, like OVN, can configure the TC datapath to
"recirculate" packets several times. If more than 4 "recirculation" rules
are configured, packets can be dropped by __tcf_classify().
Changing the maximum number of reclassifications (from 4 to 16) should be
sufficient to prevent drops in most use cases, and guard against loops at
the same time.

Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next
David S. Miller [Wed, 19 May 2021 19:58:29 +0000 (12:58 -0700)]
Merge git://git./linux/kernel/git/bpf/bpf-next

Alexei Starovoitov says:

====================
pull-request: bpf-next 2021-05-19

The following pull-request contains BPF updates for your *net-next* tree.

We've added 43 non-merge commits during the last 11 day(s) which contain
a total of 74 files changed, 3717 insertions(+), 578 deletions(-).

The main changes are:

1) syscall program type, fd array, and light skeleton, from Alexei.

2) Stop emitting static variables in skeleton, from Andrii.

3) Low level tc-bpf api, from Kumar.

4) Reduce verifier kmalloc/kfree churn, from Lorenz.
====================

3 years agoMerge branch 'mlxsw-mphash-policies'
David S. Miller [Wed, 19 May 2021 19:47:47 +0000 (12:47 -0700)]
Merge branch 'mlxsw-mphash-policies'

Ido Schimmel says:

====================
mlxsw: Add support for new multipath hash policies

This patchset adds support for two new multipath hash policies in mlxsw.

Patch #1 emits net events whenever the
net.ipv{4,6}.fib_multipath_hash_fields sysctls are changed. This allows
listeners to react to changes in the packet fields used for the
computation of the multipath hash.

Patches #2-#3 refactor the code in mlxsw that is responsible for the
configuration of the multipath hash, so that it will be easier to extend
for the two new policies.

Patch #4 adds the register fields required to support the new policies.

Patch #5-#7 add support for inner layer 3 and custom multipath hash
policies.

Tested using following forwarding selftests:

* custom_multipath_hash.sh
* gre_custom_multipath_hash.sh
* gre_inner_v4_multipath.sh
* gre_inner_v6_multipath.sh
====================

3 years agomlxsw: spectrum_router: Add support for custom multipath hash policy
Ido Schimmel [Wed, 19 May 2021 12:08:24 +0000 (15:08 +0300)]
mlxsw: spectrum_router: Add support for custom multipath hash policy

When this policy is set, only enable the packet fields that were enabled
by user space for multipath hash computation.

Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: spectrum_router: Add support for inner layer 3 multipath hash policy
Ido Schimmel [Wed, 19 May 2021 12:08:23 +0000 (15:08 +0300)]
mlxsw: spectrum_router: Add support for inner layer 3 multipath hash policy

When this policy is set, the kernel uses the inner layer 3 fields for
multipath hash computation and falls back to the outer fields if no
encapsulation was encountered. This behavior is most likely influenced
by the behavior of the flow dissector, which is used for the packet
dissection.

The Spectrum ASIC, however, cannot fallback to outer fields if inner
fields are not available. This should not result in a discrepancy from
the software data path because if several flows have matching inner
fields, they will tend to have matching outer fields as well.

Therefore, implement this policy by enabling both outer and inner layer
3 fields for the multipath hash computation.

Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: spectrum_outer: Factor out helper for common outer fields
Ido Schimmel [Wed, 19 May 2021 12:08:22 +0000 (15:08 +0300)]
mlxsw: spectrum_outer: Factor out helper for common outer fields

Outer IPv4 and IPv6 addresses are used by multiple multipath hash
policies. Factor out helpers that set these fields to increase code
sharing between different policies.

Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: reg: Add inner packet fields to RECRv2 register
Ido Schimmel [Wed, 19 May 2021 12:08:21 +0000 (15:08 +0300)]
mlxsw: reg: Add inner packet fields to RECRv2 register

The RECRv2 register is used for setting up the router's ECMP hash
configuration. Extend it with inner packet fields to allow the ECMP hash
to be calculated based on inner flow information.

Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: spectrum_router: Move multipath hash configuration to a bitmap
Ido Schimmel [Wed, 19 May 2021 12:08:20 +0000 (15:08 +0300)]
mlxsw: spectrum_router: Move multipath hash configuration to a bitmap

Currently, the multipath hash configuration is written directly to the
register payload. While this is OK for the two currently supported
policies, it is going to be hard to follow when more policies and more
packet fields are added.

Instead, set the required headers and fields in a bitmap and then dump
it to the register payload.

Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: spectrum_router: Replace if statement with a switch statement
Ido Schimmel [Wed, 19 May 2021 12:08:19 +0000 (15:08 +0300)]
mlxsw: spectrum_router: Replace if statement with a switch statement

The code was written when only two multipath hash policies were present,
so the if statement was sufficient. The next patch and future patches
are going to add support for more policies, so move to a switch
statement.

Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: Add notifications when multipath hash field change
Ido Schimmel [Wed, 19 May 2021 12:08:18 +0000 (15:08 +0300)]
net: Add notifications when multipath hash field change

In-kernel notifications are already sent when the multipath hash policy
itself changes, but not when the multipath hash fields change.

Add these notifications, so that interested listeners (e.g., switch ASIC
drivers) could perform the necessary configuration.

Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonfc: s3fwrn5: i2c: Enable optional clock from device tree
Stephan Gerhold [Wed, 19 May 2021 09:16:13 +0000 (11:16 +0200)]
nfc: s3fwrn5: i2c: Enable optional clock from device tree

S3FWRN5 depends on a clock input ("XI" pin) to function properly.
Depending on the hardware configuration this could be an always-on
oscillator or some external clock that must be explicitly enabled.

So far we assumed that the clock is always-on.
Make the driver request an (optional) clock from the device tree
and make sure the clock is running before starting S3FWRN5.

Note: S3FWRN5 asserts "GPIO2" whenever it needs the clock input to
function correctly. On some hardware configurations, GPIO2 is
connected directly to an input pin of the external clock provider
(e.g. the main PMIC of the SoC). In that case, it can automatically
AND the clock enable bit and clock request from S3FWRN5 so that
the clock is actually only enabled when needed.

It is also conceivable that on some other hardware configuration
S3FWRN5's GPIO2 might be connected as a regular GPIO input
of the SoC. In that case, follow-up patches could extend the
driver to request the GPIO, set up an interrupt and only enable
the clock when requested by S3FWRN5.

Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agodt-bindings: net: nfc: s3fwrn5: Add optional clock
Stephan Gerhold [Wed, 19 May 2021 09:16:12 +0000 (11:16 +0200)]
dt-bindings: net: nfc: s3fwrn5: Add optional clock

On some systems, S3FWRN5 depends on having an external clock enabled
to function correctly. Allow declaring that clock in the device tree.

Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonetlabel: remove unused parameter in netlbl_netlink_auditinfo()
Zheng Yejian [Wed, 19 May 2021 07:34:38 +0000 (15:34 +0800)]
netlabel: remove unused parameter in netlbl_netlink_auditinfo()

loginuid/sessionid/secid have been read from 'current' instead of struct
netlink_skb_parms, the parameter 'skb' seems no longer needed.

Fixes: c53fa1ed92cd ("netlink: kill loginuid/sessionid/sid members from struct netlink_skb_parms")
Signed-off-by: Zheng Yejian <zhengyejian1@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'intel-cleanups'
David S. Miller [Wed, 19 May 2021 19:23:25 +0000 (12:23 -0700)]
Merge branch 'intel-cleanups'

Guangbin Huang says:

====================
net: intel: some cleanups

This patchset adds some cleanups for intel e1000/e1000e ethernet driver.
====================

Reviewed-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: e1000e: fix misspell word "retreived"
Hao Chen [Wed, 19 May 2021 06:14:45 +0000 (14:14 +0800)]
net: e1000e: fix misspell word "retreived"

There is a misspell word "retreived" in comment, so fix it to "retrieved".

Signed-off-by: Hao Chen <chenhao288@hisilicon.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: e1000e: remove repeated word "slot" for netdev.c
Hao Chen [Wed, 19 May 2021 06:14:44 +0000 (14:14 +0800)]
net: e1000e: remove repeated word "slot" for netdev.c

There are double "slot" in comment, so remove the redundant one.

Signed-off-by: Hao Chen <chenhao288@hisilicon.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: e1000e: remove repeated word "the" for ich8lan.c
Hao Chen [Wed, 19 May 2021 06:14:43 +0000 (14:14 +0800)]
net: e1000e: remove repeated word "the" for ich8lan.c

There are double "the" in comment, so remove the redundant one.

Signed-off-by: Hao Chen <chenhao288@hisilicon.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: e1000: remove repeated words for e1000_hw.c
Hao Chen [Wed, 19 May 2021 06:14:42 +0000 (14:14 +0800)]
net: e1000: remove repeated words for e1000_hw.c

There are double "in" and "to" in comments, so remove the redundant one.

Signed-off-by: Hao Chen <chenhao288@hisilicon.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: e1000: remove repeated word "slot" for e1000_main.c
Hao Chen [Wed, 19 May 2021 06:14:41 +0000 (14:14 +0800)]
net: e1000: remove repeated word "slot" for e1000_main.c

There are double "slot" in comment, so remove the redundant one.

Signed-off-by: Hao Chen <chenhao288@hisilicon.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'net-dev-leading-spaces'
David S. Miller [Wed, 19 May 2021 19:17:32 +0000 (12:17 -0700)]
Merge branch 'net-dev-leading-spaces'

Hui Tang says:

====================
net: ethernet: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

        $ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
        $ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: fujitsu: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:53 +0000 (13:30 +0800)]
net: fujitsu: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Masahiro Yamada <masahiroy@kernel.org>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: 8390: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:52 +0000 (13:30 +0800)]
net: 8390: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Armin Wolf <W_Armin@gmx.de>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: xircom: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:51 +0000 (13:30 +0800)]
net: xircom: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Masahiro Yamada <masahiroy@kernel.org>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: fealnx: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:50 +0000 (13:30 +0800)]
net: fealnx: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: sun: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:49 +0000 (13:30 +0800)]
net: sun: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Vaibhav Gupta <vaibhavgupta40@gmail.com>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: smsc: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:48 +0000 (13:30 +0800)]
net: smsc: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: sis: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:47 +0000 (13:30 +0800)]
net: sis: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: "Gustavo A. R. Silva" <gustavoars@kernel.org>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: seeq: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:46 +0000 (13:30 +0800)]
net: seeq: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Masahiro Yamada <masahiroy@kernel.org>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: realtek: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:45 +0000 (13:30 +0800)]
net: realtek: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: natsemi: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:44 +0000 (13:30 +0800)]
net: natsemi: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: marvell: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:43 +0000 (13:30 +0800)]
net: marvell: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: ibm: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:42 +0000 (13:30 +0800)]
net: ibm: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'
Cc: Sukadev Bhattiprolu <sukadev@linux.ibm.com>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Acked-by: Lijun Pan <lijunp213@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: dlink: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:41 +0000 (13:30 +0800)]
net: dlink: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'
Cc: "Alexander A. Klimov" <grandmaster@al2klimov.de>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: dec: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:40 +0000 (13:30 +0800)]
net: dec: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: chelsio: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:39 +0000 (13:30 +0800)]
net: chelsio: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Rohit Maheshwari <rohitm@chelsio.com>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: broadcom: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:38 +0000 (13:30 +0800)]
net: broadcom: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: apple: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:37 +0000 (13:30 +0800)]
net: apple: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: amd: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:36 +0000 (13:30 +0800)]
net: amd: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: alteon: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:35 +0000 (13:30 +0800)]
net: alteon: remove leading spaces before tabs

There are a few leading spaces before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Jes Sorensen <jes@trained-monkey.org>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: 3com: remove leading spaces before tabs
Hui Tang [Wed, 19 May 2021 05:30:34 +0000 (13:30 +0800)]
net: 3com: remove leading spaces before tabs

There are a few leading space before tabs and remove it by running the
following commard:

$ find . -name '*.c' | xargs sed -r -i 's/^[ ]+\t/\t/'
$ find . -name '*.h' | xargs sed -r -i 's/^[ ]+\t/\t/'

Cc: Steffen Klassert <klassert@kernel.org>
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agotun: use DEVICE_ATTR_RO macro
YueHaibing [Wed, 19 May 2021 02:38:50 +0000 (10:38 +0800)]
tun: use DEVICE_ATTR_RO macro

Use DEVICE_ATTR_RO helper instead of plain DEVICE_ATTR,
which makes the code a bit shorter and easier to read.

Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoibmveth: fix kobj_to_dev.cocci warnings
YueHaibing [Wed, 19 May 2021 02:28:49 +0000 (10:28 +0800)]
ibmveth: fix kobj_to_dev.cocci warnings

Use kobj_to_dev() instead of container_of()

Generated by: scripts/coccinelle/api/kobj_to_dev.cocci

Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Lijun Pan <lijunp213@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agobpf: Make some symbols static
Pu Lehui [Wed, 19 May 2021 06:41:16 +0000 (14:41 +0800)]
bpf: Make some symbols static

The sparse tool complains as follows:

kernel/bpf/syscall.c:4567:29: warning:
 symbol 'bpf_sys_bpf_proto' was not declared. Should it be static?
kernel/bpf/syscall.c:4592:29: warning:
 symbol 'bpf_sys_close_proto' was not declared. Should it be static?

This symbol is not used outside of syscall.c, so marks it static.

Signed-off-by: Pu Lehui <pulehui@huawei.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Song Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20210519064116.240536-1-pulehui@huawei.com
3 years agobpf: Add cmd alias BPF_PROG_RUN
Alexei Starovoitov [Wed, 19 May 2021 01:40:32 +0000 (18:40 -0700)]
bpf: Add cmd alias BPF_PROG_RUN

Add BPF_PROG_RUN command as an alias to BPF_RPOG_TEST_RUN to better
indicate the full range of use cases done by the command.

Suggested-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Song Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20210519014032.20908-1-alexei.starovoitov@gmail.com
3 years agoMerge branch 'bpf-loader-progs'
Daniel Borkmann [Wed, 19 May 2021 13:27:32 +0000 (15:27 +0200)]
Merge branch 'bpf-loader-progs'

Alexei Starovoitov says:

====================
v5->v6:
- fixed issue found by bpf CI. The light skeleton generation was
  doing a dry-run of loading the program where all actual sys_bpf syscalls
  were replaced by calls into gen_loader. Turned out that search for valid
  vmlinux_btf was not stubbed out which was causing light skeleton gen
  to fail on older kernels.
- significantly reduced verbosity of gen_loader.c.
- an example trace_printk.lskel.h generated out of progs/trace_printk.c
  https://gist.github.com/4ast/774ea58f8286abac6aa8e3bf3bf3b903

v4->v5:
- addressed a bunch of minor comments from Andrii.
- the main difference is that lskel is now more robust in case of errors
  and a bit cleaner looking.

v3->v4:
- cleaned up closing of temporary FDs in case intermediate sys_bpf fails during
  execution of loader program.
- added support for rodata in the skeleton.
- enforce bpf_prog_type_syscall to be sleepable, since it needs bpf_copy_from_user
  to populate rodata map.
- converted test trace_printk to use lskel to test rodata access.
- various small bug fixes.

v2->v3: Addressed comments from Andrii and John.
- added support for setting max_entries after signature verification
  and used it in ringbuf test, since ringbuf's max_entries has to be updated
  after skeleton open() and before load(). See patch 20.
- bpf_btf_find_by_name_kind doesn't take btf_fd anymore.
  Because of that removed attach_prog_fd from bpf_prog_desc in lskel.
  Both features to be added later.
- cleaned up closing of fd==0 during loader gen by resetting fds back to -1.
- converted loader gen to use memset(&attr, cmd_specific_attr_size).
  would love to see this optimization in the rest of libbpf.
- fixed memory leak during loader_gen in case of enomem.
- support for fd_array kernel feature is added in patch 9 to have
  exhaustive testing across all selftests and then partially reverted
  in patch 15 to keep old style map_fd patching tested as well.
- since fentry_test/fexit_tests were extended with re-attach had to add
  support for per-program attach method in lskel and use it in the tests.
- cleanup closing of fds in lskel in case of partial failures.
- fixed numerous small nits.

v1->v2: Addressed comments from Al, Yonghong and Andrii.
- documented sys_close fdget/fdput requirement and non-recursion check.
- reduced internal api leaks between libbpf and bpftool.
  Now bpf_object__gen_loader() is the only new libbf api with minimal fields.
- fixed light skeleton __destroy() method to munmap and close maps and progs.
- refactored bpf_btf_find_by_name_kind to return btf_id | (btf_obj_fd << 32).
- refactored use of bpf_btf_find_by_name_kind from loader prog.
- moved auto-gen like code into skel_internal.h that is used by *.lskel.h
  It has minimal static inline bpf_load_and_run() method used by lskel.
- added lksel.h example in patch 15.
- replaced union bpf_map_prog_desc with struct bpf_map_desc and struct bpf_prog_desc.
- removed mark_feat_supported and added a patch to pass 'obj' into kernel_supports.
- added proper tracking of temporary FDs in loader prog and their cleanup via bpf_sys_close.
- rename gen_trace.c into gen_loader.c to better align the naming throughout.
- expanded number of available helpers in new prog type.
- added support for raw_tp attaching in lskel.
  lskel supports tracing and raw_tp progs now.
  It correctly loads all networking prog types too, but __attach() method is tbd.
- converted progs/test_ksyms_module.c to lskel.
- minor feedback fixes all over.

The description of V1 set is still valid:

This is a first step towards signed bpf programs and the third approach of that kind.
The first approach was to bring libbpf into the kernel as a user-mode-driver.
The second approach was to invent a new file format and let kernel execute
that format as a sequence of syscalls that create maps and load programs.
This third approach is using new type of bpf program instead of inventing file format.
1st and 2nd approaches had too many downsides comparing to this 3rd and were discarded
after months of work.

To make it work the following new concepts are introduced:
1. syscall bpf program type
A kind of bpf program that can do sys_bpf and sys_close syscalls.
It can only execute in user context.

2. FD array or FD index.
Traditionally BPF instructions are patched with FDs.
What it means that maps has to be created first and then instructions modified
which breaks signature verification if the program is signed.
Instead of patching each instruction with FD patch it with an index into array of FDs.
That makes the program signature stable if it uses maps.

3. loader program that is generated as "strace of libbpf".
When libbpf is loading bpf_file.o it does a bunch of sys_bpf() syscalls to
load BTF, create maps, populate maps and finally load programs.
Instead of actually doing the syscalls generate a trace of what libbpf
would have done and represent it as the "loader program".
The "loader program" consists of single map and single bpf program that
does those syscalls.
Executing such "loader program" via bpf_prog_test_run() command will
replay the sequence of syscalls that libbpf would have done which will result
the same maps created and programs loaded as specified in the elf file.
The "loader program" removes libelf and majority of libbpf dependency from
program loading process.

4. light skeleton
Instead of embedding the whole elf file into skeleton and using libbpf
to parse it later generate a loader program and embed it into "light skeleton".
Such skeleton can load the same set of elf files, but it doesn't need
libbpf and libelf to do that. It only needs few sys_bpf wrappers.

Future steps:
- support CO-RE in the kernel. This patch set is already too big,
  so that critical feature is left for the next step.
- generate light skeleton in golang to allow such users use BTF and
  all other features provided by libbpf
- generate light skeleton for kernel, so that bpf programs can be embeded
  in the kernel module. The UMD usage in bpf_preload will be replaced with
  such skeleton, so bpf_preload would become a standard kernel module
  without user space dependency.
- finally do the signing of the loader program.

The patches are work in progress with few rough edges.
====================

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
3 years agoselftests/bpf: Convert test trace_printk to lskel.
Alexei Starovoitov [Fri, 14 May 2021 00:36:23 +0000 (17:36 -0700)]
selftests/bpf: Convert test trace_printk to lskel.

Convert test trace_printk to light skeleton to check
rodata support in lskel.

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210514003623.28033-22-alexei.starovoitov@gmail.com
3 years agoselftests/bpf: Convert test printk to use rodata.
Alexei Starovoitov [Fri, 14 May 2021 00:36:22 +0000 (17:36 -0700)]
selftests/bpf: Convert test printk to use rodata.

Convert test trace_printk to more aggressively validate and use rodata.

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210514003623.28033-21-alexei.starovoitov@gmail.com
3 years agoselftests/bpf: Convert atomics test to light skeleton.
Alexei Starovoitov [Fri, 14 May 2021 00:36:21 +0000 (17:36 -0700)]
selftests/bpf: Convert atomics test to light skeleton.

Convert prog_tests/atomics.c to lskel.h

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210514003623.28033-20-alexei.starovoitov@gmail.com